Your other Left! Vision-Language Models Fail to Identify Relative Positions in Medical Images

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wolf, Daniel, Hillenhagen, Heiko, Taskin, Billurvan, Bäuerle, Alex, Beer, Meinrad, Götz, Michael, Ropinski, Timo
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908481859092480
author Wolf, Daniel
Hillenhagen, Heiko
Taskin, Billurvan
Bäuerle, Alex
Beer, Meinrad
Götz, Michael
Ropinski, Timo
author_facet Wolf, Daniel
Hillenhagen, Heiko
Taskin, Billurvan
Bäuerle, Alex
Beer, Meinrad
Götz, Michael
Ropinski, Timo
contents Clinical decision-making relies heavily on understanding relative positions of anatomical structures and anomalies. Therefore, for Vision-Language Models (VLMs) to be applicable in clinical practice, the ability to accurately determine relative positions on medical images is a fundamental prerequisite. Despite its importance, this capability remains highly underexplored. To address this gap, we evaluate the ability of state-of-the-art VLMs, GPT-4o, Llama3.2, Pixtral, and JanusPro, and find that all models fail at this fundamental task. Inspired by successful approaches in computer vision, we investigate whether visual prompts, such as alphanumeric or colored markers placed on anatomical structures, can enhance performance. While these markers provide moderate improvements, results remain significantly lower on medical images compared to observations made on natural images. Our evaluations suggest that, in medical imaging, VLMs rely more on prior anatomical knowledge than on actual image content for answering relative position questions, often leading to incorrect conclusions. To facilitate further research in this area, we introduce the MIRP , Medical Imaging Relative Positioning, benchmark dataset, designed to systematically evaluate the capability to identify relative positions in medical images.
format Preprint
id arxiv_https___arxiv_org_abs_2508_00549
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Your other Left! Vision-Language Models Fail to Identify Relative Positions in Medical Images
Wolf, Daniel
Hillenhagen, Heiko
Taskin, Billurvan
Bäuerle, Alex
Beer, Meinrad
Götz, Michael
Ropinski, Timo
Computer Vision and Pattern Recognition
Clinical decision-making relies heavily on understanding relative positions of anatomical structures and anomalies. Therefore, for Vision-Language Models (VLMs) to be applicable in clinical practice, the ability to accurately determine relative positions on medical images is a fundamental prerequisite. Despite its importance, this capability remains highly underexplored. To address this gap, we evaluate the ability of state-of-the-art VLMs, GPT-4o, Llama3.2, Pixtral, and JanusPro, and find that all models fail at this fundamental task. Inspired by successful approaches in computer vision, we investigate whether visual prompts, such as alphanumeric or colored markers placed on anatomical structures, can enhance performance. While these markers provide moderate improvements, results remain significantly lower on medical images compared to observations made on natural images. Our evaluations suggest that, in medical imaging, VLMs rely more on prior anatomical knowledge than on actual image content for answering relative position questions, often leading to incorrect conclusions. To facilitate further research in this area, we introduce the MIRP , Medical Imaging Relative Positioning, benchmark dataset, designed to systematically evaluate the capability to identify relative positions in medical images.
title Your other Left! Vision-Language Models Fail to Identify Relative Positions in Medical Images
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.00549